Colour matters: the effects of lensing on the positional offsets between optical and submillimetre galaxies in Herschel-ATLAS

TL;DR

This study reveals that lensing effects influence the positional offsets between Herschel-ATLAS submillimetre sources and their optical counterparts, especially for redder and brighter sources, impacting source identification and redshift estimation.

Contribution

The paper demonstrates that weak lensing significantly affects the positional offset distributions in submm surveys, a factor previously underappreciated in counterpart identification.

Findings

Redder and brighter sources show broader offset distributions.

Weak lensing by foreground structures explains the offset variations.

Implications for redshift and luminosity function estimates.

Abstract

We report an unexpected variation in the positional offset distributions between Herschel-ATLAS sub-millimetre (submm) sources and their optical associations, depending on both 250-{\mu}m signal-to-noise ratio and 250/350-{\mu}m colour. We show that redder and brighter submm sources have optical associations with a broader distribution of positional offsets than would be expected if these offsets were due to random positional errors in the source extraction. The observation can be explained by two possible effects: either red submm sources trace a more clustered population than blue ones, and their positional errors are increased by confusion; or red submm sources are generally at high redshifts and are frequently associated with low-redshift lensing structures which are identified as false counterparts. We perform various analyses of the data, including the multiplicity of optical associations, the redshift and magnitude distributions in H-ATLAS in comparison to HerMES, and simulations of weak lensing, and we conclude that the effects are most likely to be explained by widespread weak lensing of Herschel-SPIRE sources by foreground structures. This has important consequences for counterpart identification and derived redshift distributions and luminosity functions of submm surveys.